fresh-2-layer-medmcqa-distill-of-bert-base-uncased-gpqa

This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 9.2527
• Accuracy: 0.4242

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0005
• train_batch_size: 16
• eval_batch_size: 16
• seed: 321
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	63	9.8453	0.2626
No log	2.0	126	12.6958	0.3131
No log	3.0	189	10.7690	0.3232
No log	4.0	252	9.5414	0.3737
No log	5.0	315	9.9080	0.3687
No log	6.0	378	9.5831	0.3889
No log	7.0	441	9.5607	0.3687
2.7217	8.0	504	10.5312	0.3283
2.7217	9.0	567	9.4693	0.4040
2.7217	10.0	630	9.5568	0.3889
2.7217	11.0	693	9.0092	0.3636
2.7217	12.0	756	8.9660	0.3889
2.7217	13.0	819	9.2727	0.3838
2.7217	14.0	882	9.3829	0.3636
2.7217	15.0	945	8.9537	0.3889
0.4611	16.0	1008	9.1312	0.3939
0.4611	17.0	1071	9.2527	0.4242
0.4611	18.0	1134	9.2069	0.4242
0.4611	19.0	1197	9.0783	0.4091
0.4611	20.0	1260	9.0630	0.4040

Framework versions

• Transformers 4.34.0.dev0
• Pytorch 2.0.1+cu117
• Datasets 2.14.5
• Tokenizers 0.14.0